T2 Systems Canada Inc.
Author: Tim Cummins
'To use machine learning responsibly, there is a need to ensure values are aligned.' This comment by a representative of Google sums up a key dilemma with all relationships - no matter whether the intelligence being applied is human or machine-based. In the world of business, 'mismatched objectives' (or expectations) lie at the heart of many disputes.
Contracts exist in large part because of these mismatches. In theory, a good contracting process serves two purposes – one is to reduce the chances of misalignment, the other is to deal with its consequences. These are demanding concepts – and contracts are not always good at dealing with them. How could we make them better?
That is a question which goes to the heart of IACCM's purpose and its research has consistently pointed to answers (and the underlying causes). Among the issues / solutions:
– organisational measurement and reward systems: these typically do not offer incentives that ensure 'matched objectives'. They should be changed, especially for those involved in designing and negotiating contracts.
– attitudes to risk: for all the talk about risk, the focus of terms and conditions remains weighted towards areas such as liabilities, indemnities, intellectual property – not on safeguarding that objectives are – and remain – aligned. Again, this approach comes from custom and choice.
– coherent governance: change is ever-present and increasingly rapid, yet for many the approach to its management has not changed. Fears of 'scope creep' or challenges in budgeting result in failure to use the right form of contract (e.g. agile, relational) and to develop agreed change forums and methods.
Ironically, it may require the discipline of machine programming to overcome these deep-seated problems. One benefit from automation is that it is not subject to the ingrained habits of humans!